CN-121999009-A - Distributed multi-target fusion tracking method and device based on time calibration

CN121999009ACN 121999009 ACN121999009 ACN 121999009ACN-121999009-A

Abstract

The invention provides a distributed multi-target fusion tracking method and device based on time calibration, and relates to the technical field of sensors. The distributed multi-target fusion tracking method based on time calibration comprises the steps of constructing a posterior probability density expression of time calibration parameters by utilizing a Gaussian mixture approximation technology and a probability hypothesis density filter based on a generalized covariance cross fusion criterion, selecting the mean value of the Gaussian component with the largest weight from a plurality of Gaussian components of the posterior probability density expression as an estimated value of the time calibration parameters, and fusing a first multi-target probability density function and a calibrated second multi-target probability density function based on the generalized covariance cross fusion criterion to obtain a multi-target state estimated value of a tracking target. The invention can efficiently and accurately fuse the data from a plurality of sensor nodes in a distributed multi-target tracking scene so as to realize the accurate positioning of the tracking target.

Inventors

LI GUCHONG
WANG XUEQIAN
LI GANG
HE YOU

Assignees

西北工业大学
清华大学
清华大学深圳国际研究生院

Dates

Publication Date: 20260508
Application Date: 20251229

Claims (10)

1. The distributed multi-target fusion tracking method based on time calibration is characterized by comprising the following steps of: Based on a generalized covariance cross fusion criterion, a Gaussian mixture approximation technology and a probability hypothesis density filter are utilized to construct a posterior probability density expression of a time calibration parameter, wherein the posterior probability density expression is in a Gaussian mixture exponential function form, and the time calibration parameter is used for realizing the alignment of a first multi-objective probability density function and a second multi-objective probability density function in a time dimension; Selecting the mean value of the gaussian component with the largest weight from a plurality of gaussian components of the posterior probability density expression as the estimated value of the time calibration parameter; according to the estimated value of the time calibration parameter, the second multi-objective probability density function is aligned with the first multi-objective probability density function in time, and a calibrated second multi-objective probability density function is obtained; And based on a generalized covariance cross fusion criterion, fusing the first multi-target probability density function and the calibrated second multi-target probability density function to obtain a multi-target state estimation value of the tracking target.
2. The distributed multi-objective fusion tracking method based on time calibration according to claim 1, wherein the constructing a posterior probability density expression of time calibration parameters based on generalized covariance cross fusion criteria using gaussian mixture approximation technique and probability hypothesis density filter comprises: Determining an initial posterior probability density expression of the time calibration parameter by using a generalized covariance cross fusion criterion, wherein the initial posterior probability density expression of the time calibration parameter comprises a prior probability density function of the time calibration parameter and likelihood functions of the time calibration parameter between the first multi-objective probability density function and the second multi-objective probability density function; Determining a Gaussian mixture exponential form of the likelihood function by using a Gaussian mixture approximation technology and a probability hypothesis density filter; and obtaining a posterior probability density expression of the time calibration parameter in the Gaussian mixture index form according to the likelihood function in the Gaussian mixture index form and the prior probability density function.
3. The time-alignment-based distributed multi-objective fusion tracking method of claim 2, wherein said determining a gaussian mixture exponential form of the likelihood function using a gaussian mixture approximation technique and a probability hypothesis density filter comprises: Determining a mathematical expression of the likelihood function using a probability hypothesis density filter; The mathematical expression of the likelihood function is expressed in the form of a gaussian mixture index using a gaussian mixture approximation technique.
4. A distributed multi-objective fusion tracking method based on time alignment according to claim 3, wherein said determining a mathematical expression of said likelihood function using a probability hypothesis density filter comprises: Processing the first multi-objective probability density function by using a probability hypothesis density filter to obtain a first intensity function; processing the second multi-objective probability density function by using a probability hypothesis density filter to obtain a second intensity function; and using the first intensity function and the second intensity function to represent the likelihood function, and obtaining a mathematical expression of the likelihood function.
5. The time-alignment-based distributed multi-objective fusion tracking method of claim 4, wherein the likelihood function is expressed using the following formula: ; Wherein, the ; Wherein, the As a function of the first intensity of the light, For a second intensity function under the time calibration parameters, State variables for tracking the target; is a variable related to a time alignment parameter; Is a weight coefficient; The gaussian mixture exponential form of the likelihood function is formulated as follows: ; Wherein a and b are index numbers respectively, For the number of gaussian components corresponding to the first multi-objective probability density function, For the number of gaussian components corresponding to the second multi-objective probability density function, For the weight coefficients, for balancing the contributions of the different terms to the objective function, Is in the form of normal distribution, In order to calibrate the parameters in time, the time-alignment parameters, Is the mean value of the gaussian component, Is the variance of the gaussian component.
6. The distributed multi-objective fusion tracking method based on time calibration according to claim 1, wherein the fusing the first multi-objective probability density function and the calibrated multi-objective probability density function of the second probability hypothesis density filter based on the generalized covariance cross fusion criterion to obtain a multi-objective state estimation value of the tracking objective comprises: based on a generalized covariance cross fusion criterion, fusing the first multi-target probability density function and the calibrated second multi-target probability density function to obtain a Gaussian component associated with a multi-target state of the tracking target; and selecting the average value of Gaussian components with the weight value larger than a preset threshold value from the Gaussian components associated with the multi-target state as the multi-target state estimation value of the tracking target.
7. A distributed multi-target fusion tracking device based on time alignment, comprising: The construction module is used for constructing a posterior probability density expression of a time calibration parameter based on a generalized covariance cross fusion criterion by utilizing a Gaussian mixture approximation technology and a probability hypothesis density filter, wherein the posterior probability density expression is in a Gaussian mixture exponential function form, and the time calibration parameter is used for realizing the alignment of a first multi-objective probability density function and a second multi-objective probability density function in a time dimension; A selection module, configured to select, from a plurality of gaussian components of the posterior probability density expression, a mean value of a gaussian component with a largest weight as an estimated value of the time calibration parameter; the synchronization module is used for aligning the second multi-objective probability density function with the first multi-objective probability density function in time according to the estimated value of the time calibration parameter to obtain a calibrated second multi-objective probability density function; and the fusion module is used for fusing the first multi-target probability density function and the calibrated second multi-target probability density function based on a generalized covariance cross fusion criterion to obtain a multi-target state estimation value of the tracking target.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and running on the processor, wherein the processor implements the time-alignment-based distributed multi-objective fusion tracking method of any of claims 1 to 6 when the computer program is executed by the processor.
9. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor implements a distributed multi-target fusion tracking method based on time alignment according to any of claims 1 to 6.
10. A computer program product comprising a computer program which, when executed by a processor, implements a distributed multi-objective fusion tracking method based on time alignment according to any of claims 1 to 6.

Description

Distributed multi-target fusion tracking method and device based on time calibration Technical Field The invention relates to the technical field of sensors, in particular to a distributed multi-target fusion tracking method and device based on time calibration. Background The distributed multi-target tracking technology has key effects in the fields of military reconnaissance, intelligent transportation, unmanned aerial vehicle coordination and the like, realizes stable tracking of a dynamic target by fusing multi-sensor node data, but with the scale expansion and isomerism enhancement of a sensor network, a tracking system faces multiple challenges. The time asynchronism problem causes inconsistent sampling period, initial time difference and transmission/processing time delay among sensor nodes, data are difficult to align in the time dimension (such as different sampling frequencies of a radar and a camera or disordered wireless transmission), and the continuity of a target track is damaged. The existing centralized synchronization method relies on a global clock or frequent calibration, is high in calculation and communication cost and is difficult to meet the real-time requirement. The data association ambiguity is derived from remarkable difference of observation characteristics of heterogeneous sensors (radar, infrared and vision) on the same target, and noise interference, so that cross-modal data (such as point cloud and image) is difficult to match, in the related technology, target mistracking or loss often occurs due to mismatching of characteristic dimensions or fuzzy association rules, the calculation efficiency bottleneck is that all node data need to be transmitted to a central processor due to a centralized architecture, the data quantity increases exponentially along with the number of nodes, so that processing delay is caused, and the complexity of a traditional algorithm is extremely high under a multi-target scene, so that the traditional algorithm is difficult to adapt to real-time requirements of a high dynamic environment. Therefore, how to efficiently and accurately fuse data from multiple sensor nodes in a distributed multi-target tracking scene to realize accurate positioning of a tracking target is a technical problem to be solved. Disclosure of Invention The invention provides a distributed multi-target fusion tracking method and device based on time calibration, which are used for solving the defects in the prior art, and the data from a plurality of sensor nodes are fused efficiently and accurately in a distributed multi-target tracking scene so as to realize accurate positioning of a tracking target. The invention provides a distributed multi-target fusion tracking method based on time calibration, which comprises the following steps of. The method comprises the steps of constructing a posterior probability density expression of a time calibration parameter based on a generalized covariance cross fusion criterion by using a Gaussian mixture approximation technology and a probability hypothesis density filter, wherein the posterior probability density expression is in a Gaussian mixture exponential function form, the time calibration parameter is used for realizing alignment of a first multi-objective probability density function and a second multi-objective probability density function in a time dimension, the first multi-objective probability density function is obtained by filtering data collected by a reference sensor, the second multi-objective probability density function is obtained by filtering data collected by a sensor to be registered, a mean value of a Gaussian component with the largest weight is selected from a plurality of Gaussian components of the posterior probability density expression to serve as an estimated value of the time calibration parameter, the second multi-objective probability density function is aligned with the first multi-objective probability density function according to the estimated value of the time calibration parameter, and a calibrated second multi-objective probability density function is obtained by fusing the first multi-objective probability density function and the calibrated second multi-objective probability density function based on the generalized covariance cross fusion criterion. According to the distributed multi-target fusion tracking method based on time calibration, the posterior probability density expression of the time calibration parameter is constructed by using Gaussian mixture approximation technology and probability hypothesis density filter based on generalized covariance cross fusion criterion, and the method comprises the following steps: The method comprises the steps of determining an initial posterior probability density expression of a time calibration parameter by using a generalized covariance cross fusion criterion, determining a Gaussian mixture index form of the likelihood function by using a Gaussia